[GRASS-user] question on r.watershed updates in GRASS 6.4.RC3

[Markus Metz]

On Tue, May 14, 2013 at 9:26 AM, Hanan Karam
<HananKaram at esinternational.com> wrote:
> Hi Markus,
>
> I'm wondering about the nature of the recent bug fixes to r.watershed listed for the 6.4.RC3 release (sync skipping edge cells; fix bug in SFD).    I have some results obtained using r.watershed in 6.4.RC2.  Do you think these would have been affected by the bugs listed above, and would you recommend re-running my scripts using the updated module?

These were small bugs, only relevant for elevation models with NULL
cells. The differences should be very small.

> Also, do you know if r.watershed in 6.4.RC3 and in a recent Grass 7 version for Windows(r56229) are consistent?

Not for SFD. In GRASS 6.x, surface flow is still creeping along edges
(region borders), this has been changed only in GRASS 7. MFD should be
identical. r.watershed in GRASS 7 is faster, in particular for the
memory mode which reduces memory consumption and keeps intermediate
data on disk.

Markus M